Fragmenting Linux is not the way to beat Apple

In an attempt to beat the iPhone, more competitors are developing their own versions of Linux. This is a mistake.

In an attempt to copycat Apple's hardware-plus-software vertical approach to the mobile market, the Linux industry is fragmenting fast and risks undermining its best chance for beating the iPhone.

The mobile Linux market has always had more variants/distributions than sense, ranging from Google Android to LiMo to Moblin (now MeeGo) to Bada to WebOS name it. Whereas Linux has been a rallying force in the enterprise server market, with diverse competitors and partners collaborating on a common code base to save costs and boost innovation, in the mobile market Linux has tended toward entropy.

Such entropy is about to get worse, even as signs abound that things should be improving. First, Hewlett-Packard announced its acquisition of Palm to "double down on WebOS" development and thereby "compete aggressively in the market with Apple and Google."

Next, rumors are swirling that Motorola has bought Azingo, another mobile Linux vendor. Motorola has been a big proponent of LiMo, an attempt to corral different vendors around a common mobile Linux code base, and has also been a primary distributor of Android.

If Motorola acquires Azingo, will this end? Some speculate that Azingo is merely a way to develop an Android alternative for the China market, but this doesn't answer the question of fragmentation that such a move raises.

Meanwhile, Intel and Nokia have fused together their Moblin and Maemo projects under the MeeGo brand, while Samsung backs Bada and Google stands atop the heap with Android.

If only they could all get together on this and focus on a common Linux foundation, then compete on top of it with their different user environments, as I presented at the CELF Embedded Linux Conference 2010 in San Francisco a month ago:

A partial solution to a fragmented mobile Linux market? Matt Asay

Again, this is now the norm for Linux servers, with companies as diverse as Red Hat, Canonical, Texas Instruments, Google, Intel, IBM, Oracle, etc. all working furiously to build a great core and then competing in the packaging, hardware, etc. that surrounds it.

But this isn't what we're getting in mobile Linux. In mobile we have a resurrection of the failed Unix wars, with competing vendors owning their own operating systems. Google Android is not compatible with MeeGo...or Azingo...or LiMo...or get the point.

It's possible that the fragmentation will resolve itself. Android, according to new data from NPD, has just surpassed the iPhone in quarterly unit shipments:


Android may become the de facto mobile Linux out of sheer dominance.

But there's still a growing perception that to compete with Apple one must own the OS, which may push competing vendors to buy rival Linux distributions rather than collaborate. Motorola co-CEO Sanjay Jha said as much on Motorola's first-quarter earnings call:

I've always felt that owning your OS is important, provided you have an ecosystem, you have all the services and you have an ability and the scale to execute on keeping that OS at the leading edge. And I continue to believe that at some point, if we have all of those attributes, that owning our own OS will be a very important thing.

It's understandable why Apple's competitors would want to compete with Apple on its own terms. It's also foolish. To beat Apple, the industry needs to collaborate, even as it has done in servers to fix the Unix mess and meet the growing Microsoft Windows threat. Buying an OS is not the solution.

It may well be the problem.